Poverty is a problem faced both nationally and globally. Central Sulawesi Province is one of the provinces in Indonesia with an economic growth rate that is above the national rate, which is inversely proportional to the amount of poverty that occurs. This study aims to find a model using ordinal logistic regression analysis to predict predictor variables that affect poverty and rank the influencing predictor variables using dominant analysis. Based on 11 predictor variables that are thought to be closely related to the characteristics of poverty with the approach of poor households in Central Sulawesi province, ordinal logistic regression analysis concludes that seven predictor variables that affect it are X1 (floor area), X4 (roof), X5 (latrine facilities), X6 (drinking water), X9 (credit), X10 (assets), and X11 (aid rice). The importance of the seven predictor variables can be ranked using dominant analysis. The dominant analysis concludes that the highest importance in influencing poverty in Central Sulawesi province is the variable X11 (aid rice) followed by variables X6 (drinking water), x5 (latrine facilities), X1 (floor area), X4 (roof), X10 (assets), and X9 (credit).
30 June 2020 in Jurnal Aplikasi Statistika & Komputasi Statistik
Sulawesi Tengah adalah salah satu Provinsi di Indonesia yang memiliki permasalahan dalam perceraian. Tingkat perceraian di Sulawesi Tengah pada tahun 2016 sebesar 2,44%. Persentase tingkat perceraian di Sulawesi Tengah ini menjadi tingkat perceraian ketiga tertinggi di Indonesia. Pada penelitian ini diteliti faktor-faktor yang mempengaruhi kasus perceraian di Sulawesi Tengah. Metode yang digunakan adalah regresi probit biner dengan variabel respon adalah status perkawinan. Hasil penelitian menunjukkan bahwa variabel prediktor yang mempengaruhi perceraian secara signifikan di Provinsi Sulawesi Tengah adalah umur kawin pertama (X2) kategori 1 (18-21 tahun) dan kategori 2 ( >21 tahun), tingkat pendidikan (X3) kategori 1 (SD) dan kategori 4 (di atas SMA), daerah tempat tinggal (X4) kategori 1 (kota) dan jumlah pengeluaran rumah tangga (X6) dengan tingkat ketepatan klasifikasi model sebesar 99,2%.
1 April 2025 in Bulletin of Electrical Engineering and Informatics
Due to limited space and regulations, food labels often lack information on micronutrients, i.e., vitamins and minerals. Accurately predicting missing these micronutrient data is essential yet challenging. This study explores the feasibility of using machine learning to predict these missing nutrients based on a limited reported nutrient (protein and carbs). Using the Tabel Komposisi Pangan Indonesia (TKPI) dataset, we evaluated the performance of 12 diverse classifiers to predict binary classes ("low" or "high") for 13 target micronutrients. Random forest emerged as the best performing classifier with an average accuracy of 0.7421 across all target nutrients. Additionally, we introduced feature engineering techniques by incorporating polynomial and ratio features to enhance model performance. Minimum redundancy maximum relevance (mRMR) feature selection was then applied to identify the most informative features. This approach boosted the average accuracy of the random forest classifier to 0.7591. These findings highlight the efficacy of feature engineering and selection in enhancing nutrient prediction models, demonstrating the potential to improve consumer knowledge about unknown nutrients in food.
Background/Objectives: Glioblastoma multiforme (GBM) is an aggressive and heterogeneous brain tumor with poor prognosis, emphasizing the need for reliable molecular biomarkers to improve patient stratification and treatment planning. This study aimed to identify key genes associated with overall survival in GBM by employing and comparing machine learning (ML) and deep learning (DL) approaches using RNA-Seq gene expression data. Methods: RNA-Seq expression and clinical data for primary GBM tumors were obtained from The Cancer Genome Atlas (TCGA). A univariate Cox proportional hazards regression was used to identify survival-associated genes. For survival prediction, ML-based feature selection techniques—RF, GB, SVM-RFE, RF-RFE, and PCA—were used to construct multivariate Cox models. Separately, DeepSurv, a DL-based survival model, was trained using the significant genes from the univariate analysis. Gradient-based importance scoring was applied to determine key genes from the DeepSurv model. Results: Univariate analysis yielded 694 survival-associated genes. The best ML-based Cox model (RF-RFE with 90% training data) achieved a c-index of 0.725. In comparison, DeepSurv demonstrated superior performance with a c-index of 0.822. The top 10 genes were identified from the DeepSurv analysis, including CMTR1, GMPR, and PPY. Kaplan–Meier survival curves confirmed their prognostic significance, and network analysis highlighted their roles in processes such as purine metabolism, RNA processing, and neuroendocrine signaling. Conclusions: This study demonstrates the effectiveness of combining ML and DL models to identify prognostic gene expression biomarkers in GBM, with DeepSurv providing higher predictive accuracy. The findings offer valuable insights into GBM biology and highlight candidate biomarkers for further validation and therapeutic development.